Preparing Posthumanist Perspectives on AI-Human Collaboration in Developing Cyber Ethics Curricula

Preparing Posthumanist Perspectives on AI-Human Collaboration in Developing Cyber Ethics Curricula

Straight - Lowery - Poehlman - Yowika

University of Arizona

Introductions

Ryan Straight, PhD
Assistant Professor
ryanstraight@arizona.edu

Jonathon Hernandez-Lowery
Graduate
jrl84623@arizona.edu

David Poehlman
Undergraduate
dpoehlman@arizona.edu

Waamene Yowika
Undergraduate
waamene@arizona.edu

Follow along!

Since the Paper

AI-generated case studies in CYBV 329
Focus on student engagement & outcomes
25 completed survey responses
Mixed-methods assessment

Key Findings

Strong student engagement
High effectiveness ratings
Positive learning outcomes
Rich qualitative feedback

First, apologies for not being there in person. I was meant to be but my travel plans were completely upended three different times. So, here we are!

Now, a quick little recap of the paper that my students and I put together: we collaborated on the development of AI-generated scenarios or case studies to use in an class on Cyber Ethics, Law, and Policy. Research has shown that the use of realistic but fabricated case studies is highly effective and we wanted to try applying generative AI to this. Turns out, it works quite well!

I asked my in-person class this fall (after having gone through 5 different case studies, 4 fabricated and 1 real) about their experiences and thoughts on the fabricated ones. That’s what I want to share in this presentation, mainly, as it will drive a full study and paper to follow.

So, the entire class responded, which was great. And here’s the first of what we found. The vast majority of students believed that the case study and its analysis hit the learning objectives for the module and effectively hit the topics they were on. I’ll be sharing some of these shortly.

Study Background

AI-generated case studies in cyber ethics
Novel pedagogical approach
Collaboration between human & AI
Focus on student engagement

Key Innovation

Realistic scenarios for complex topics
Integration with existing curricula
Balanced technical & ethical content

This work represents a shift from traditional human-centric approaches by embracing AI not just as a tool, but as a collaborative agent in developing educational materials. Our informal pilot study with 25 students revealed fascinating insights about this human-AI pedagogical partnership. This isn’t just about using AI to create content–anybody can do that these days–rather, it’s about recognizing AI as a co-participant in the educational process, what Katherine Hayles might call a ‘cognitive assemblage’ between human and machine.

Looking at our innovation metrics on the right, we see particularly high scores in scenario development and ethics coverage - over 85% effectiveness in both areas. This isn’t coincidental; it reflects how the human-AI collaboration could address complex ethical scenarios from multiple perspectives simultaneously. The bar chart shows two other key innovation areas: student feedback (overall feelings of innovation) and learning integration. Notice how the integration scores, while still strong at 78%, are lower - this represents the ongoing challenge of seamlessly incorporating these posthuman approaches into traditional educational frameworks.

Student Engagement

72% strongly agreed with effectiveness
High interest in case study format
Deep engagement with materials

Notable Elements

Real-world scenarios
Contemporary issues
Practical applications

Here we have a rather compelling story about student engagement with this posthuman pedagogical approach. Looking at the darkest blue segments at the top of each bar, we see strong agreement across three key measures: capturing interest, engagement versus traditional lectures, and motivation for deep thinking. Over 72% reported high engagement, indicating strong acceptance of this posthuman pedagogical approach. What’s particularly interesting is how the ‘More Engaging than Lectures’ category shows the highest proportion of strong agreement - suggesting students may actually prefer learning from materials that emerge from human-AI collaboration rather than purely human instruction. This aligns with what philosophers like Karen Barad describe as ‘intra-action,’ where meaning emerges through the relationship between human and technological actors.

To drive this home, I genuinely had students asking questions like, “Wait, is this real?”

Learning Outcomes

Enhanced understanding of concepts
Increased confidence in application
Better grasp of complexities

Student Feedback

More practical than traditional methods
Clear connection to real-world issues
Strong theoretical foundation

Development Process

Module Design
- 7 core modules, 5 case studies, 4 fabricated
- Integration with NSA requirements
- Focus on practical application
AI Collaboration
- Mostly Anthropic’s Claude implementation
- Iterative refinement process
- Student feedback integration

graph TD
    A[Course & NSA Objectives] --> B[AI Generation]
    B --> C[Student Review]
    C --> D[Refinement]
    D --> E[Implementation]
    E --> F[Feedback]
    F --> D
    style A fill:#f9f,stroke:#333
    style B fill:#bbf,stroke:#333
    style C fill:#dfd,stroke:#333

Process details to highlight:

Each module carefully aligned with NSA knowledge requirements to match that Cyber Ethics knowledge unit
AI generates initial scenarios based on learning objectives and integrate the readings
Student collaborators provide crucial feedback
Iterative refinement ensures quality and relevance
Continuous improvement based on implementation feedback

Here we see our development methodology visualized as a Mermaid flowchart, illustrating what Karen Barad calls ‘intra-action’ in practice. Starting with course objectives, moving through AI generation, student review, refinement, implementation, and feeding back into refinement - this isn’t a linear process but a dynamic assemblage of human and non-human actors. The feedback loop represents how both human and AI contributions evolved through iteration. Of particular note is how we used Anthropic’s Claude models for scenario generation - not as a tool to be wielded, but as a ‘companion species’ in the creative process. This again drives home that the scenarios weren’t simply human-authored cases about technology but emerged through genuine technological co-creation.

Scenario Structure

Background documents
Character profiles
Social media elements
Timeline progression

Integration

Real-world parallels
Multiple perspectives
Ethical considerations
Policy implications

Important elements to emphasize:

Each scenario includes multiple types of documents and perspectives
Characters represent different stakeholders in cyber ethics
Social media elements add realism and contemporary relevance
Timeline allows for development of complex ethical situations

The bar chart on the right shows the relative importance of different scenario components, with documents and ethics scoring highest at 95% and 94% respectively. This structure reflects what Braidotti calls ‘transversal connections’ - crossing traditional boundaries between technical and social elements. Each scenario includes background documents, character profiles, social media elements, and timeline progression, creating what I call a ‘posthuman narrative assemblage.’ Looking at the purple bars, you can see how policy and ethics components rank nearly as high as technical documentation - this balanced integration is crucial for posthuman pedagogy where no single perspective dominates.

Topic Coverage

Ethics: Strong foundation
Policy: Clear understanding
Law: Practical application

Integration Success

Balanced coverage
Interconnected topics
Real-world relevance

EXODUS

CO(G)NECT

BELLATOR

Common Themes in Student Feedback

Real-world Application
- Industry relevance
- Current issues
Learning Experience
- Engaging format
- Clear objectives
Critical Thinking
- Complex scenarios
- Multiple perspectives

Recommendations

Expand Interactive Elements
- More hands-on components
- Group discussions
Development Areas
- Scenario complexity
- Timeline management
- Additional topics
Research
- Long-term impact
- Comparative analysis

Future Directions

Comparative effectiveness studies
Long-term impact assessment
Integration with other disciplines

Next Steps

Expanded scenario library
Enhanced interactive elements
Cross-institutional implementation

Thank you!

Participate in a Study! Visit mavrxlab.org/pphp for details.

ryanstraight.com

mavrxlab.org

linkedin.com/in/ryanmstraight